2
\$\begingroup\$

i have avalanche photo diode with which i am trying to digitalize the current from TIA through a comparator, the diode biased using a HVDCDC

the source is being a 20ns laser with a PRF of 3Khz

the circuit is as below

schematic

simulate this circuit – Schematic created using CircuitLab

enter image description here

the output of opamp for an high intensity light is like above, we can see the opamp fully saturated, but i am clue less about the over shoot and ringing effect, this happened when i have put my bias voltage 300V, i am not able to stop the opamp to enter in to saturation.

even if i see the datasheet of DC-DC CA05N, it can maximum source 2mA of current, and the peak current of didoe is only 0.25mA, so i understood the dcdc current is exceeding 2mA

so to avoid the opamp going in to saturation i tried to use a 300K in series with the source but still the opamp went to saturation,

it evident that in order to make the opamp not to go in to saturation, the only way i can try is to reduce the bias voltage to as less as possible, so i operated it at 70V for which the response is as below.

enter image description here

reasons are unknow for the expansion of 20ns pulse to 100ns even through the APD is having a BW of 70MHz,

so in order to operate the APD properly, i understood i have to bias with higher voltage to see proper response for low light, but at the same time i have to avoid saturation.

i feel limiting the current to the TIA will solve problem, kindly suggest some ways to bypass the excess current which is to the opamp.

Kindly throw some light on behaviour of APD in saturation also

EDIT:

as the answers suggested to go for a unity gain stable equivalent, i went for changing my opamp to LTC6269 which is 400Mhz unity gain stable

below is the result i got with photo diode polarity inversed, so you may see a inverted output, but effect persists

enter image description here

the analog support was suggesting me also on using a subber at the photo diode bias so i used a 20ohm in series with 100p cap thinking this disturbance is due to the power supply i am using, frankly i am clueless how to tweak a snubber

please help me in identifying the problems rootcause

**EDIT2:**i have to tweak my Cf accordingly i feel, to avoid this, i dont think this is ringing, i feel i have solved my problem , will comeback with results shortly

EDIT3 : replacing Cf with 4pF given me a clean pulse with out any ringing Thanks for the help

EDIT4: But after few experiments i identified that i am not saturating the photodiode/opamp completely, so i have increase the pulse width and pulse power on photo diode with which my response turned like below

enter image description here

\$\endgroup\$
3
\$\begingroup\$

The amplification and reaction speed of an APD depend on the current through it. It's a bad habit of the physics community to use APDs with a voltage source and a series resistor for biasing. As you have witnessed, this fails in high dynamic range applications. I suggest you change your biasing circuit to provide a constant current instead of constant voltage.

The ringing you see is from your opamp. You are using a variant that is not unity gain stable. The -10 in the part number denotes that it's a decompensated opamp for a gain of at least 10 (Linear's opamp numbering convention). But for this kind of circuit to work, you need unity gain stability.

I suggest you have a look at Linear Appnote 92 which discusses a few circuits for APDs.

\$\endgroup\$
  • \$\begingroup\$ i understood that i should have constant current source in place, but also limited to current less than my saturation current of opamp.can you please given how to provide such constant current source based biasing for my application. i am unable to find such app notes for voltages high like 300V, also how can i modify my exiting ckt to keep opamp stable, i feel i have to keep my gain as low as 10 \$\endgroup\$ – kakeh Oct 1 '17 at 4:53
  • \$\begingroup\$ any how to keep it stable i am maintaining cin / cf ratio to 10 \$\endgroup\$ – kakeh Oct 1 '17 at 5:10
  • \$\begingroup\$ i have repeated the experiment using a stable opamp to see the effect but the phenomenon persists, kindly help \$\endgroup\$ – kakeh Nov 15 '17 at 9:14
  • \$\begingroup\$ I guess it's time to open up a new question and show us the new circuit. \$\endgroup\$ – Attila Kinali Nov 19 '17 at 17:40
  • \$\begingroup\$ i mistook the parasitic ringing, i have put a valid 4pF cap in feed back because which the stable opamp did not pose any problems, but its still a unsolved question how to exploit the full bandwidth of the decompensated opamps, recently LT rolled out a 10GHz GBWP opamp, what are they for if they are not useful? \$\endgroup\$ – kakeh Nov 20 '17 at 9:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.